Reinforcement Learning for Mobile Robotics Exploration: A Survey.

Area of Science:

• Robotics and Artificial Intelligence
• Machine Learning for Autonomous Systems

Background:

• Autonomous mobile robots require efficient exploration of unknown environments.
• Integrating reinforcement learning (RL) with robotics is a growing research area for developing advanced exploration strategies.

Purpose of the Study:

• To provide a comprehensive review of recent research on RL-based exploration strategies for single and multi-robot systems.
• To analyze the current state of the art, facilitating future research in this interdisciplinary domain.

Main Methods:

• Systematic review of academic literature linking reinforcement learning and robotic exploration.
• Analysis of employed RL algorithms, their integration into exploration strategies, and common challenges addressed.

Main Results:

• Compilation and analysis of RL algorithms used in robotic exploration.
• Identification of how robotic exploration addresses key RL challenges (exploration-exploitation, dimensionality, reward shaping, convergence).
• Overview of experimental setups, software tools, and progress achieved.

Conclusions:

• Reinforcement learning offers promising solutions for autonomous robot exploration.
• Further research is needed to address limitations and enhance the robustness of RL-based exploration strategies in complex scenarios.